skip to main content
Merck Manual LogoMERCK MANUALConsumer Version
Search icon

Survey Shows More Than 3 in 5 Americans Use AI Tools for Medical Information, but Most Take Steps to Check Sources and Information

Commentary04/13/26 The Manual's Editorial Staff

Many Americans are turning to AI tools for medical information, but there are conflicting views about the trustworthiness of the results. More than three in five (62%) Americans have used AI tools like medical chatbots, ChatGPT, Google Assistant, Siri, or Google Gemini for medical information, according to a recent poll conducted online by The Harris Poll on behalf of Merck Manuals among 2,057 U.S. adults. About a third of Americans (32%) say they do not trust the medical information provided by AI tools. Nine in 10 who’ve used AI for medical info say they take steps to check the legitimacy of information provided by AI.

“Artificial intelligence is a powerful tool for accessing and organizing information across all sorts of tasks, and finding health information is no exception,” said Dr. Sandy Falk, M.D., Merck Manuals Editor-in-Chief. “One of AI’s great capabilities is as a search engine that finds not only links to medical content but also puts together information from a variety of sources. The challenge is making sure the information is accurate and reliable and provided in the right context. At the Merck Manuals, we’re committed to empowering people to use trusted medical information to make the best choices for their health, whether that’s through our own AI search tool or by powering other AI applications. Our mission is to provide resources patients and health care professionals can use to ensure AI results are correct or to follow up an AI answer with a deeper dive.”

The Merck Manuals/Harris Poll survey found Americans use AI tools to access medical information in a variety of ways, with the most common being:

  • Getting information about a specific condition or disease (33%)
  • Getting more information about symptoms they are experiencing (29%)
  • Getting advice about nutrition, lifestyle habits, and physical activity (26%)
  • Asking about medication side effects or dosage (26%)

Among Americans who use AI tools for medical information, some common actions they take when using include:

  • Entering a list of symptoms to get a diagnosis (54%)
  • Entering detailed prompts into AI that include things such as vitals, medical history or other important personal health data (44%)
  • Using an AI tool specifically designed to focus on health and medical topics (39%)

Among those who use AI for medical information, nearly 9 in 10 (88%) say they trust the information, but only 14% say they trust it completely. This trust among people who use AI for medical information was relatively consistent across generations (87% each of Gen Z ages 18-29 and Millennials ages 30-45, 90% of Gen X ages 46-61, 88% of Baby Boomers ages 62-80).

Nine in 10 Americans who use AI tools for medical information (90%) say they do take steps to check the legitimacy of the information provided by AI. The most common tactics include:

  • Talk to a healthcare professional for confirmation or clarification (41%)
  • Cross-reference with other reputable AI or online information platforms (e.g., search engines, websites) (39%)
  • Check the sources provided within the AI response for accuracy and/or authority (37%)
  • Use academic and medical databases for deeper research (32%)

Different demographics are more likely to use AI for medical information. Gen Z and Millennials are more likely to use AI for medical information than Gen X and Baby Boomers, while Baby Boomers are the least likely to do so (77% each Gen Z and Millennials, 58% Gen X, 40% Baby Boomers). Parents with children under 18 are more likely to use AI for medical information, with more than 8 in 10 (83%) saying they use it compared to just 52% of those who are not parents with children under 18. The most common uses among parents of minors are to get information about a specific condition or disease (46%) and to get more information about symptoms they were experiencing (41%).

 

Consumer Survey Methodology

This survey was conducted online within the United States by The Harris Poll on behalf of Merck Manuals from February 26-March 2, 2026, among 2,057 U.S. adults ages 18 and older, among whom 1,285 use AI for medical information. The sampling precision of Harris online polls is measured by using a Bayesian credible interval.  For this study, the sample data is accurate to within +/- 2.7 percentage points using a 95% confidence level. This credible interval will be wider among subsets of the surveyed population of interest. For complete survey methodology, including weighting variables and subgroup sample sizes, please contact kmiller@gobraithwaite.com.